Restructuring 'for AI' is this cycle's digital transformation. Most will fail for the same reason

From AI layoffs to operating model pivots: the strategic test

AI-driven restructuring and new workforce operating models are becoming the latest corporate mantra. Senior leaders now frame almost every large workforce transformation as a shift toward an AI-first enterprise design, yet the pattern looks uncomfortably close to the last digital transformation wave. The same promises about technology, data, and talent are back, but the same structural mistakes in organizational design and workforce planning are already visible.

The numbers around the workforce are stark and should reset how every business thinks about this moment. According to Challenger, Gray & Christmas’ midyear 2024 report, AI-attributed job cuts reached 101,743 through June, nearly double the 54,836 recorded in full-year 2023, and artificial intelligence has led all layoff reasons for four consecutive months. When AI becomes the top stated reason for change, the burden of proof on leaders and their strategy for human work, skills, and employee engagement rises dramatically.

Two very different AI restructuring stories illustrate the fork in the road for organizations. In early 2024, GitLab cut 14% of its workforce, exited 22 countries, and reorganized research and development into roughly 60 autonomous units while deploying AI agents for internal reviews and approvals, and revenue still grew 23% in the same quarter with gross margin and net retention holding steady, as detailed in its Q1 FY2025 shareholder letter. Cloudflare, in its 2024 SEC filings and earnings commentary, cut about 20% of its organization and declared an agentic, AI-first operating model, with the CEO stating that AI and agents are now core parts of the workforce and of how teams, roles, and decision making will operate in real time.

These moves sound similar on the surface, but the underlying operating models and workforce strategies differ in ways that will shape future work. The strategic test is simple and unforgiving for leaders and boards. Does the restructuring come with reinvestment in new human–machine capabilities, skill development, and data strategy, or does headcount simply drop to the bottom line while business outcomes quietly erode over the next 24 months?

When AI-enabled restructuring and workforce redesign are real, you see a clear capability model for the post-restructuring organization. You see explicit workforce planning that maps which knowledge workers move into new roles, which tasks shift to artificial intelligence, and which teams are redesigned around human–machine collaboration. You also see change management that treats people as assets to redeploy, not costs to remove, and that is where most organizations are still at the foundation stage.

When the strategy is cosmetic, the signals are equally clear to any CHRO. There is no explicit operating model blueprint, no quantified data strategy, and no plan for continuous learning that would make the workforce future ready. In those cases, AI is a label on a cost program, and the same leaders will be back in two years explaining why productivity, innovation, and employee engagement all stalled despite the promise of transformation.

The anatomy of a genuine AI era operating model

Executives who lived through the last digital transformation cycle remember how many projects were justified as strategic but delivered only short-term cost savings. The same risk now surrounds every AI-enabled workforce redesign, and the only antidote is a much more rigorous operating model design. A genuine AI era playbook starts before any workforce change with a hard audit of how work actually gets done across the organization.

That audit must map the full system of work, not just the org chart. It should trace how data flows through processes, how technology supports or constrains teams, and how human judgment shapes decision making at each step. In practice, this means sitting with knowledge workers, frontline managers, and leaders to understand which tasks are repeatable, which require human nuance, and where artificial intelligence can augment rather than replace human work.

From there, CHROs and COOs need a capability model for the future work design, not just a headcount model. The operating model should specify which capabilities sit with humans, which sit with AI agents, and which are shared in human–machine workflows that operate in real time. It should also define the new roles that orchestrate these systems, such as AI product owners, data strategy leads, and workforce transformation architects who can translate business outcomes into concrete operating models.

GitLab’s restructuring offers one of the clearest examples of this capability-first mindset. By reorganizing research and development into roughly 60 autonomous units, the company effectively created smaller operating cells where AI agents and humans can co-evolve work patterns. The CEO explicitly framed the move as agentic era positioning, not pure cost cutting, and committed to reinvesting the vast majority of savings into new technology, talent, and skill development, including expanded AI features in the product and internal enablement programs.

Contrast that with restructurings where AI is invoked but the operating model remains opaque. When a company announces a 20% workforce reduction and an AI-first strategy without publishing a clear capability map, the odds are high that the change is primarily financial. The Oracle restructuring in 2024, analyzed by market commentators as a capital expenditure driven move rather than a productivity play, is a cautionary tale for leaders tempted to copy headline numbers without understanding the underlying business context.

For CHROs, the practical implication is blunt. Do not sign off on any AI-powered workforce transformation that cannot show, on one page, how the new operating model will improve specific business outcomes such as cycle time, error rates, or revenue per employee. If the only quantified benefit is reduced payroll, you are not leading a workforce transformation, you are managing a budget cut with better branding.

Real operating models also embed governance from day one. Gartner’s 2024 forecasts on agentic AI adoption expect that over 40% of agentic AI projects will be canceled within a few years due to inadequate governance and unclear business value, a projection based on early project failure rates and stalled pilots. A robust governance design clarifies who owns AI decisions, how data is validated, how bias is monitored, and how change management will support employees as their work, skills, and roles evolve.

Workforce architecture, not headcount math: the CHRO’s critical path

Most AI-related restructuring decks still start with a spreadsheet, not a workforce architecture. That is the core reason this cycle risks repeating the failures of digital transformation, where technology spend outpaced human capability building by a wide margin. A serious workforce transformation begins with a granular map of roles, skills, and work outcomes, then layers AI on top of that reality.

Workforce architecture means understanding the true composition of your workforce, not just by job title but by task, skill, and value contribution. It requires segmenting knowledge workers, frontline employees, and specialist talent into clusters based on the type of human work they perform and the degree to which artificial intelligence can augment or automate it. This is where a disciplined data strategy becomes a strategic asset rather than a compliance exercise.

In practice, CHROs should insist on three artefacts before any AI-related restructuring is approved. First, a task-level inventory that shows which activities are candidates for automation, which are candidates for augmentation, and which must remain fully human due to risk, ethics, or customer expectations. Second, a workforce planning model that simulates different operating models over a three- to five-year horizon, including redeployment paths, skill development requirements, and financial-services-style scenario analysis for risk.

Third, a transition plan that goes far beyond severance and basic change management. A credible plan includes continuous learning pathways, internal talent marketplaces, and explicit commitments to retrain at least a portion of affected employees into future ready roles. It also includes clear metrics for employee engagement, retention, and internal mobility, because those are the leading indicators of whether the organization is actually building a sustainable human–machine operating model.

To make this tangible, CHROs can use a simple one-page capability map as a working template:

Regulation is quietly raising the bar for this kind of discipline. The European Union’s HR AI rules, and the extended compliance reprieve analyzed in recent guidance for CHROs on AI compliance, effectively force organizations to treat AI systems as part of the workforce. That means documenting how AI influences decision making in hiring, promotion, and performance, and it pushes leaders to design operating models where accountability for outcomes remains clearly human.

For US-based organizations, the regulatory pressure may feel distant, but the operating implications are immediate. If your AI-enabled operating model cannot withstand the scrutiny of a European-style audit of data, algorithms, and human oversight, it is probably too fragile to support long-term business outcomes. The same rigor that regulators demand is the rigor that protects your organization from failed projects and reputational damage.

One global financial-services firm offers a concrete illustration of this shift from headcount math to workforce architecture. Facing margin pressure in 2023, the company initially modeled a 15% reduction in operations roles through automation of claims processing. After building a task-level inventory, leaders realized that many employees held critical tacit knowledge about fraud patterns and exception handling. Instead of a pure reduction, they moved to a dual-track plan: roughly half of the targeted roles were redeployed into new positions such as AI workflow designers and fraud analytics specialists, supported by a structured reskilling program. Within 18 months, the firm reported lower claim cycle times, improved fraud detection rates, and higher internal mobility, while keeping overall labor costs in line with the original target.

CHROs should also revisit how they present workforce transformation to the board. Instead of leading with layoff numbers, lead with the new capability model, the human–machine design, and the investment in skill development and continuous learning. Boards are increasingly attuned to the risk that short-term savings from workforce cuts can undermine long-term strategy, and they will respond to a narrative that connects AI, operating models, and sustainable value creation.

From cost cutting to capability compounding: a quarterly playbook for leaders

The most important shift for leaders is to treat AI-enabled restructuring and operating model change as a compounding capability bet, not a one-time cost event. That requires a different cadence of decisions, metrics, and governance than traditional restructuring programs. It also demands that CHROs, CFOs, and CIOs operate as a single strategic team rather than as separate functional silos.

Over the next quarter, executives can run a focused operating model sprint instead of a sprawling transformation program. Start by selecting two or three critical value streams where AI can realistically change how work is done, such as customer onboarding, claims processing, or software release management. For each value stream, map the current human work, the supporting technology, the data flows, and the decision making points, then design a target state where human–machine collaboration improves speed, quality, and employee engagement.

Next, define explicit workforce planning scenarios for those value streams. Scenario A might emphasize automation, reducing headcount but risking loss of tacit knowledge and weakening the organization’s ability to adapt to change. Scenario B might emphasize augmentation, keeping more talent in place but requiring heavier investment in skill development, continuous learning, and new roles such as AI coaches or workflow designers who sit inside teams.

Leaders should then choose based on measurable business outcomes, not on narrative appeal. If automation delivers a 10% cost reduction but degrades customer satisfaction and increases error rates, while augmentation delivers a smaller cost benefit but improves both metrics, the strategic choice is obvious. This is where a disciplined data strategy, with real-time dashboards and clear KPIs, turns AI rhetoric into operating reality.

Governance must keep pace with this experimentation. The same Gartner 2024 projection that over 40% of agentic AI projects will be canceled due to weak governance is a direct warning to organizations that launch pilots without clear ownership. A simple but powerful rule is that every AI system affecting people or customers must have a named human owner, a documented risk assessment, and a quarterly review of outcomes, including unintended effects on the workforce and on organizational culture.

To help CHROs and their peers apply this discipline, a simple flow-style checklist can anchor each quarterly review:

Finally, leaders need to rethink how they signal value to employees living through these shifts. Rituals around tenure, progression, and recognition are changing in the future work landscape, as explored in recent analysis of how to mark a five-year anniversary in the new world of employment. In an AI-reshaped organization, the most powerful signal is not a plaque or a bonus, but a visible path into new roles, new skills, and meaningful human work that technology cannot easily replace.

Restructuring for AI can either hollow out the workforce or make it future ready. The difference lies in whether leaders treat operating models as living systems that blend human judgment, artificial intelligence, and robust data into better decisions. In this cycle, the winning organizations will be those that optimize not for engagement scores, but for stay signals.

Key figures on AI restructurings and operating models

• AI-attributed job cuts reached 101,743 through June 2024, nearly double the 54,836 recorded in full-year 2023, making artificial intelligence the leading stated reason for layoffs for four consecutive months according to Challenger, Gray & Christmas data.
• GitLab reduced its workforce by 14%, exiting 22 countries and reorganizing research and development into roughly 60 autonomous units, while still reporting 23% revenue growth in the same quarter and signaling reinvestment into AI capabilities in its Q1 FY2025 materials, illustrating how an AI era operating model can pair cost savings with capability building.
• Cloudflare cut approximately 20% of its employees and declared an agentic AI-first operating model in its 2024 SEC filing and earnings commentary, with the CEO stating that AI and agents are now core parts of the workforce and of how the business operates, including real-time decision support.
• Gartner’s 2024 outlook on agentic AI projects projects that over 40% of such initiatives will be canceled within a few years due to inadequate governance and unclear business value, highlighting the risk of launching AI programs without a robust operating model and workforce strategy.
• In many large organizations, labor represents between 50% and 70% of operating expenses, which means that any AI-enabled restructuring that focuses solely on headcount reduction without capability building risks undermining long-term business outcomes.